(Not) finding phonetics that have Outlier entries in Pleco

[Gharial]

I know Outlier's stuff is a work in progress, and that its coverage is therefore going to be a bit patchy still and potentially with minor errors or fudges etc, but I have to say the integration of its entries so far into Pleco is pretty disappointing in terms of easily accessing phonetic components that have any allied Outlier info.

That is, some "Outlierized" phonetic components can only be found by searching for the PC itself rather than whatever fuller character it may form part of (fair enough, provided the student knows how to parse things and despite the apparent lack of any guidance); or worse, can only be found as part of another character (characters? For those who luck out more than once, or aspire to be exhaustive above and beyond), which introduces a truly unacceptable level of pot-luck ferreting (the ostensibly easy jump-around functionality of Pleco notwithstanding).

For example, if you search for 拗 or similarly-structured characters you will currently draw a blank for Outlier FORM info etc, but if you search for 幼 itself it's about the only result. Conversely, 奄 will draw a blank for Outlier info as it's only available (as far as I can tell) via 掩. Another example would be 矣, found only at 埃 (and one has no way of really telling that from the CHAR Compounds tab without clicking on and viewing each individual result).

I think this Outlier add-on could've done with a fair bit more work by Outlier and testing by Pleco before release, or was the content really so unique that the Chinese-learning world couldn't possibly live another minute without it being released sooner than not. I'm definitely going to be wary of upgrading to the Expert edition (if and when that's ever "ready").

[mikelove]

I’ll leave most of this to Outlier, but it seems like you basically want a way to search for characters that don’t have their own entries yet but are described in other entries, correct?

I think the plan is to fill in entries for most of those, but if that’s not imminent it would not be too difficult for us to generate stub entries that pointed you to the entries containing that character via links.

[Gharial]

Correct Mike, and I'd imagine ANYONE who bought this add-on was expecting better integration and functionality, but has been too polite or something to (openly) say so.

 

So, if you can now do the actually not-so-difficult work of generating stubs or better, then yes, please do! People might even thank you for it.

 

From a purely logical point of view it would've made more sense for Outlier (especially given the way Pleco failed to integrate things) to complete entries like 奄 before 掩, but there presumably was a good reason for them doing some the "wrong" way around, eh?

[mikelove]

I’m not sure how you would have liked us to integrate it better; it’s a character dictionary and we let you look up entries in it by character.

 

It would generally not make a whole lot of sense to bring up 掩 when you search for 奄 except in a “characters containing component” tab, and we already offer one of those. Showing people 掩 in the search result list when you enter 奄 would seem like a bug. It makes sense right now to make that link via a stub entry because 奄 lacks a standalone entry, but that’s a temporary data issue, not an integration one.

[Gharial]

"I’m not sure how you would have liked us to integrate it better; it’s a character dictionary and we let you look up entries in it by character."

 

"Hey look, a character dictionary (despite all the Outlier hullabaloo about components), with half its entries missing and no dependable indexing for the other half! Wow!"

 

Hmm perhaps you could've been honest in the marketing? That might've been helpful. "Unless the exact item you're interested in already has a dedicated Outlier entry, you won't be able to find it easily if at all, even if it's a component in another Outlier entry, simply because Pleco can't be bothered to do the necessary linking work and fully harness the power of digital devices in this particular instance, at least not until Outlier has finished everything and sent it by pack donkey via Outer Mongolia. We're feeling as stiffed as you poor customers. No, really, we are".

 

Actually I'm not too bothered (certainly not in discussing the utility of the CHAR > Components function for searching pre-existing dictionaries' entries, nor in the exact way Pleco might implement a temporary fix for the Outlier, that's obviously your job! Should you deign to do so. And the main problem had you read that bit correctly is of not being able to get to Outlier info on 奄 from just 奄 rather than currently and necessarily from 掩 - the latter is obviously too roundabout~dependent on lucking out), as the patchiness of Outlier's entries has already got me looking elsewhere (not that I'm ultimately THAT interested in the minutiae of paleography, though it's obviously now a trend that has to be at least noted if not "fully" catered for), but I'd be surprised if this hasn't been an issue for any other users (certainly those more dependent on possibly just Pleco), hence my feeling compelled to raise it here. (Call me irascible but I do find it quite irritating when reference works for Chinese neglect or mess up ease of access to whatever information they might contain).

[mikelove]

The search for 奄 should link to the dedicated entry for it, that’s my point. Linking to the entry for a character containing it is not ideal, if for no other reason than because it’s not really tailored to explaining that component in general but only to explaining it in the context of that character.

 

The information you’re so concerned about not being able to find is merely a short definition “to spread out wide in order to cover s.t.” which is not substantially different from that in other Pleco dictionaries, along a statement that it functions as both a sound and a meaning component, which of course is only interesting in the context of it being the entry for 掩.

 

So this is a feature that would make little or no logical sense in the finished dictionary; if 奄 had its own entry it would be useless and even confusing. I can understand why you might want it now to fill in the gaps (albeit only a little bit), but I’m not prepared to accept blame for the fact that we didn’t originally adapt the dictionary in a way that assumed this sort of workaround would be necessary / desirable.

[Gharial]

You can continue making your "point" as often as you like Mike, but in my opinion Pleco dropped the ball on this one. If the user is not directed from searching for e.g. 奄 (just one random example, if that component really is of no possible import in itself) to Outlier information for it (currently only within an Outlier entry for 掩) then like it or not Pleco is not being as maximally useful as it could be, and it is disingenuous of you to pretend otherwise. If it had been made clear that the Outlier Essentials (I'm certainly not being suckered into buying the Expert) wouldn't be fully functional until all relevant entries (uniquely literal entry points) were complete then I doubt many would have bought it. They'd've correctly surmised that accessing any Outlier information could be a problem.

 

Putting all that another way, Pleco's apparent obstinacy here is compounding the lack (slow release etc) of Outlier's entries in general.

[mikelove]

OK, obviously I disagree with you on that, but since it seems like there's not much I can do to convince you otherwise I'll stop here.

[Gharial]

How does the option of refunding people sound? Does that work for Pleco? Cos that's the way it might be going.

 

You see, I've got no money left because I spent it all on the Pro version of Pleco and then a very limited Outlier add-on. The Pro package isn't bad, but the Outlier isn't functional enough for my tastes, but hey, I'm just a customer whose views about ease of use obviously count for little. I should have read far more into the term 'character dictionary' than is fair or sensible for an otherwise digital e-product, and apparently now be able to mindread under precisely which fuller character Outlier may or may not have so far dealt with any component of potential interest. And of course I will not be falling back on Karlgren or Harbaugh or whatever else because they really do have nothing dependable to say, not even about how to better or more logically organize a work dealing with the evolution of Chinese characters.

 

Next thing you'll be telling us is that there may be some components that could remain out of easy reach until one invests in only the largest Outlier offering, fully forthcoming...2025ish? One thing I did manage to look up though was the ancient Chinese equivalent of 'locked in', which simply had a picture of Outlier and Pleco shaking hands and cackling in lieu of a proper entry.

 

I know, I know, I should just launch a dictionary project on Kickstarter and take years to deliver anything, but I'm not comfortable with even the idea of doing that.

[mikelove]

If possible, we're always happy to offer refunds if customers are dissatisfied, with this or any other add-on.

 

For purchases made through pleco.com or Google Play, we can generally refund purchases within 1 year, as long as your credit card is still valid.

 

For purchases made through the App Store, I'm afraid that your refund request would have to go through Apple - developers are not able to initiate refunds on App Store purchases ourselves.

[Gharial]

I paid via those Google Play scratchcard things. You know what though, thanks for the offer but it's fine, forget it, Outlier probably really needs the money and are doing partway valuable work it seems. It's just a shame they're so slow and Pleco hasn't quite run with the ball yet (again, just IMHO). Anyway, I've done what I can to raise what I think is an issue or at least a potential concern. Either way, have a nice day Mr Love.

[Ash@Outlier]

@Gharial

Mike (and Pleco) certainly did NOT drop the ball here. They've only done what we asked them to do. We did not request the functionality you're asking for. As to us being slow, yes, we are slow. We're slow because I have to do all of the computer programming required to make tools to do the work faster AND I do the paleography. So, if I'm spending time writing a software tool to make the process go more quickly, then paleography has to stop and wait on that. To date, I've created tools to speed up converting paleographic analyses into dictionary entries, creation of the meaning trees, and creation of the expert entries, not to mention the software for data testing, fixing Python 2's issue with extension characters, creating data lists, etc. etc. Each one of those has taken a significant amount of time to code.

 

Our internal rule for making an entry for a given component is "does it show up more than once in the most common 2000 characters as something other than a sound component" OR "is it a common character itself?" In the most common 2000 characters, 奄 shows up in only 掩. Given that I've been in graduate school for 12 years in Taiwan using Chinese as the medium of instruction and don't recall ever having come across 掩 in my reading, I doubt the frequency stats on it are even correct. When I first starting learning Chinese, one of the first words I learned to write is 鉛筆. I lived in Taiwan for over 10 years before coming across 鉛 in reading. So, this idea of putting out data for the most common characters first is not just an abstract idea to me. 

 

FYI: I did do the analysis on 奄 before doing the analysis on 掩. It just didn't make it into the dictionary because getting common characters / components done is a far higher priority (and it should be). Here are my notes on it:

to spread out wide in order to cover s.t.； 廣大申展所覆蓋 ；从大从申，會廣大申展所覆蓋之意。按：奄 *ʔ(r)omʔ、申 *l̥i[n]、大*lˤa[t]-s     西周    HDK    4015
HDK = 黃德寬《古文字譜系疏證》. If you have questions that are bothering you this much, it would be better for everyone concerned if you just joined our Facebook Group and asked them directly ("Chinese Character Discussion" group, if you're interested) or email me at ash at outlier-linguistics dot com.

 

BTW, the having the Expert version is not a requirement for getting updates on the Essentials version. Anyone that buys the Essentials version will get updates on it for as long as there are updates, system-level data included. I've finished writing the software tool for speeding up the Expert entries, but am working on getting a version that John can also use. This has taken a lot more time than I originally thought (surprise, surprise). I've already started making Expert entries with the new tools. Once John has a version he can use, then we'll put out 100 new Expert entries as fast as we can make them. I know you said you aren't interested in buying it. That's fine. We'll be doing it regardless.

Edited October 21, 2019 at 04:15 AM by Ash@Outlier
to make the meaning clearer

[imron]

2 minutes ago, Ash@Outlier said:

and don't recall ever having come across 掩 in my reading

It's not uncommon in 武侠 and you also see it in names of martial arts moves, e.g. there is a move in some Taichi styles called 掩手宏锤. 

[Ash@Outlier]

Good to know. That also explains why I haven't seen it. : )

[Gharial]

Quote

We did not request the functionality you're asking for.

 

Well, well done you! And if you want to use frequency as such an arbiter (than merely one factor to note if and when building things up more logically), then cut out even more entries, I mean, who'd complain, right? Certainly not Pleco, it seems. Customers, who are they? People to ignore and condescend to with remarks like "I’m not sure how you would have liked us to integrate it better; it’s a character dictionary and we let you look up entries in it by character" (but what about the all-important components?!) and now in your huffy rambling post.

 

I know I don't seem appreciative enough, but you guys couldn't even find an organizing principle for your lightweight components poster until I suggested simply putting the feckin' things in English keyword alphabetical order to help sidestep the supposedly huge huge issue of any resemblance to radicals (no idea if you implemented that, don't care frankly), so please forgive my not taking the whole Outlier approach too seriously now. Good luck with your workload, it sounds truly awful and completely thankless. Perhaps hire an extra staff member or two (no, I'm not available LOL), or if that would impact profits too much, maybe an intern you can push around.

 

As for the fabled updates ("system-level data included") when will those be exactly? And will you need to twist Mike's arm a bit to get him to actually do the necessary work at Pleco's end? It's obviously a slow day here at my new start-up Outliar as for some reason I haven't been able to find the items I'm interested in looking up.

[imron]

For someone who objects so strongly to condescension and huffy rambling posts, you sure do enough of it yourself  ¯\_(ツ)_/¯.

[Gharial]

If Pleco and Outlier were more receptive here to customer concerns and suggestions (or *GASP* had actually ANTICIPATED them) then the tone might be a bit more positive. (If you concentrate REALLY hard then you can just about see where things went rapidly downhill. Partly I confess in my second post, though I see that got a like even so, but certainly from Mike's second). And I note that Ash couldn't resist popping up to "correct" me despite my having signed off to Mike in a reasonably polite manner and considering the discussion (such as it is) over.

 

But OK, now we're onto the surefire topic of who's the greatest condescending huffy poster, sure, maybe I could challenge Ash for that crown. I think he's the winner though when it comes to whining about how unappreciated one feels.

[imron]

20 minutes ago, Gharial said:

If you concentrate REALLY hard then you can just about see where things really went downhill

Reading through the thread, all I see mike and Ash being polite to a customer who is being increasingly rude and obnoxious because he doesn't like the answers he's getting, and who takes offense when none was given.

 

I don't see a problem with Mike's second post.  He was certainly more diplomatic that I would have been based on your second post.  Maybe your third post could have simply asked if it would be possible to add a 'component' search feature instead of being full of snark, sarcasm and veiled insinuations of dishonesty.

[Gharial]

Yes, nothing untoward, unhelpful or arrogant about Outlier turning out to have specifically requested the otherwise standard and expected functionality NOT be added, end of, not ours to question why. Must've saved Pleco some work in the short term, but could lose or alienate a few customers in the long run.

 

I think the tone of my responses is actually quite justified given all the above. Next time I'll know better than to offer any feedback - that, or myself drink enough of the Kool-Aid that I don't have to. Cheers all round, undisturbed status quo restored. It's nice to be an unvalued customer.

[vellocet]

Dude, you're being a complete dick about a software function that doesn't work the way you think it should work.  

 

Quote

Correct Mike, and I'd imagine ANYONE who bought this add-on was expecting better integration and functionality, but has been too polite or something to (openly) say so.

I bought it.  I think we're lucky to have anything like it at all, and complaining it isn't as feature-laden as you'd like it is really rich.  It shouldn't even exist, from what I understand someone stole the money raised by Kickstarter and that would have been the end of it except it's someone's academic work.  You really have to make a distinction between professional polished products and works of narrow interest created by enthusiastic amateurs.